Agent♥︎Age
Catalog

sqlite-graph-memory

Official

by tonydzi · Python

Recall over your markdown notes: vector search plus your [[wikilinks]] as a graph, in SQLite.

The io.github.tonydzi/sqlite-graph-memory MCP server provides “recall over your markdown notes” by combining vector search with [[wikilinks]] interpreted as a graph, using SQLite. It is described as storing and querying note content to support graph-based retrieval alongside semantic search.

🛠️ Key Features

  • Vector search over markdown notes
  • Graph model built from [[wikilinks]]
  • SQLite-backed storage/querying

🚀 Use Cases

  • Retrieving information from a personal markdown note collection
  • Navigating linked note relationships via wikilinks
  • Finding relevant notes using both semantics (vectors) and connections (graph)

⚡ Developer Benefits

  • Works with markdown notes and wikilinks as the primary inputs
  • Uses a single SQLite datastore for the recall workflow

⚠️ Limitations

  • The available details describe vector search and wikilink graph recall, without further specification of tool surface area.

Topics

agent-memoryclaude-codegraph-ragllm-agentsobsidianragsqliteai-agentsknowledge-graphmcpmcp-serversecond-brainvector-searchwikilinksclaude-memorycodexcursorpersistent-memorykarpathy-llm-wikillm-wiki